Prospective Evaluation of Effect of Metformin on Activation of AMP-activated Protein Kinase (AMPK) and Disease Control in a Sub-group Analysis of Patients with GI Malignancies

Background: Observational studies have demonstrated association of metformin with reduced cancer incidence and mortality in multiple cancer types, including gastrointestinal (GI) malignancies. Anti-neoplastic effects of metformin are believed through many mechanisms including activation of AMP-activated protein kinase, which controls mammalian target of rapamycin (mTOR) growth regulatory pathway.

Methods: In a pilot, delayed-start randomized study, non-diabetic patients with GI cancers were randomized to 2 arms, Stage 1: concurrent metformin (500mg twice daily) plus chemotherapy vs. chemotherapy alone followed by cross over to metformin plus chemotherapy arm in Stage 2, while adverse events (DLT) were assessed by CTCAE v.3.0. As a translational correlate, we used phosphorylation of AMPKα at Thr172 to measure AMPK activation by western blot technique in PBMCs isolated from patients before and after receiving M. These levels were correlated with radiological (RECIST 1.1) and tumor marker outcomes by descriptive analysis. In this study, we present the sub-group analysis of patients with GI cancers.

Results: 41 patients with GI cancers (colorectal: 22, pancreatic: 12, gastroesophageal: 4, biliary: 2, others: 1) were treated in this trial. Mean duration of metformin therapy was 85 days (range: 9-443). There was no significant difference in grade 3 or above DLT in metformin plus chemotherapy vs. chemotherapy arm (14% vs. 12% respectively). Gel band density analysis on 19 patients showed that 63% patients had increased phosphorylation of AMPKα after metformin (ratio of phospho-AMPKα after and before metformin > 1) with mean = 1.227 (± 0.134). RECIST 1.1 restaging showed disease control in 55% patients and 45% patients had decline in tumor markers. Of note, 60% of patients with disease control also showed increase in phosphorylation of AMKα.

Conclusions: This group of patients treated with metformin prospectively demonstrates the impact of metformin on AMPKα phosphorylation, and correlates with clinical benefit in patients with GI cancers when metformin was added to systemic chemotherapy of varying types. We aim to perform a dose-escalation of metformin in our next study with additional metabolomics correlates.